 |
 |
 |
 |
|
Many formerly used Department of Defense (DoD) bombing ranges are now in the process of being turned over to the public domain. Prior to opening these lands to the public, it is necessary to survey these sites to ensure that any unexploded ordnance (UXO) has been removed from the site. Over the past 10 years, the DoD has invested in the development of geophysical sensors for the efficient subsurface characterization of these sites. However, the large size of the sites (10's of thousands of acres) and the relatively high expense of exhaustive site characterization has prompted the need for statistically-based approaches to UXO site characterization. This project is funded by the Strategic Environmental Research and Development Program (SERDP) to develop a statistically-based sampling protocol to be used at UXO sites.
The official title of this project is "Bayesian Approach to UXO Site Characterization with Incorporation of Geophysical Information". This project is examining ways to map the occurrence of UXO on DoD sites using geostatistical estimation algorithms. The intent is to develop a characterization protocol that will use geostatistics to accurately estimate the intensity of UXO at unsampled locations. Several technical issues are being addressed to effectively utilize geostatistical estimation. These issues include kriging in finite domains (transect data), cokriging of scrap metal and UXO and Bayesian incorporation and updating of geophysically derived prior estimates of UXO.
Papers:
- Saito, H., P. Goovaerts and S.A. McKenna, (in review), Impact of Accuracy of Prior Information and Sensors on Geostatistical Characterization of Unexploded Ordnance (UXO) Sites, submitted to the UXO/Countermine Forum, Sept. 3rd-6th, 2002, Orlando, Florida
- Bilisoly, R.L., S.A. McKenna and E. Cespedes, (in review) Spatial Filtering of Magnetometer Data for Improved UXO Discrimination, submitted to the UXO/Countermine Forum, Sept. 3rd-6th, 2002, Orlando, Florida
- McKenna, S.A., H. Saito and P. Goovaerts, (in review), Estimating the Spatial Distribution of UXO from Limited Sample Data Using Geostatistics, submitted to the UXO/Countermine Forum, Sept. 3rd-6th, 2002, Orlando, Florida
- Saito, H., P. Goovaerts and S.A. McKenna, (in review), Combining Logistic Regression with Kriging for Mapping the Risk of Occurrence of Unexploded Ordnance (UXO), submitted to: IAMG Annual Meeting, Berlin, Germany, September, 2001.
- McKenna, S.A., Application of a Doubly Stochastic Poisson Model to the Spatial Prediction of Unexploded Ordnance, invited paper, accepted for presentation at 2001 International Association of Mathematical Geology annual meeting, September, 9-12, 2001, Cancun, Mexico.
|
|
|
